Add new SentenceTransformer model
Browse files- 1_Pooling/config.json +10 -0
- README.md +775 -0
- config.json +23 -0
- config_sentence_transformers.json +14 -0
- model.safetensors +3 -0
- modules.json +14 -0
- sentence_bert_config.json +4 -0
- special_tokens_map.json +51 -0
- tokenizer.json +0 -0
- tokenizer_config.json +66 -0
- vocab.txt +0 -0
1_Pooling/config.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"word_embedding_dimension": 768,
|
3 |
+
"pooling_mode_cls_token": false,
|
4 |
+
"pooling_mode_mean_tokens": true,
|
5 |
+
"pooling_mode_max_tokens": false,
|
6 |
+
"pooling_mode_mean_sqrt_len_tokens": false,
|
7 |
+
"pooling_mode_weightedmean_tokens": false,
|
8 |
+
"pooling_mode_lasttoken": false,
|
9 |
+
"include_prompt": true
|
10 |
+
}
|
README.md
ADDED
@@ -0,0 +1,775 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- sentence-transformers
|
4 |
+
- sentence-similarity
|
5 |
+
- feature-extraction
|
6 |
+
- dense
|
7 |
+
- generated_from_trainer
|
8 |
+
- dataset_size:80
|
9 |
+
- loss:MultipleNegativesRankingLoss
|
10 |
+
base_model: microsoft/mpnet-base
|
11 |
+
widget:
|
12 |
+
- source_sentence: How many different active substances were detected in surface water
|
13 |
+
across all catchment areas?
|
14 |
+
sentences:
|
15 |
+
- 'metabolites were not detected in the water bodies.
|
16 |
+
|
17 |
+
2.1.1. Antibiotics/Enzyme-Inhibitors and
|
18 |
+
|
19 |
+
Abacavir in Surface-Water
|
20 |
+
|
21 |
+
Fifty detections were found in all catchment areas in surface water, which corresponds
|
22 |
+
to 15 different active substances:
|
23 |
+
|
24 |
+
12 antibiotics, two enzyme inhibitors, and one antiviral. The number of detections
|
25 |
+
per sampling station ranged from 0 to 7
|
26 |
+
|
27 |
+
different active substances. The Ave river-Prazins (Santo Tirso) and Serzedelo
|
28 |
+
I and II (Guimar ã es) as well as Ria
|
29 |
+
|
30 |
+
Formosa-coastal water (Faro and Olh ã o), each one with two sampling sites, showed
|
31 |
+
the most detected compounds in'
|
32 |
+
- '2. Results
|
33 |
+
|
34 |
+
2.1. Frequency of Detections:
|
35 |
+
|
36 |
+
Antibiotics/Enzyme-Inhibitors and Abacavir
|
37 |
+
|
38 |
+
in Surface-Groundwater
|
39 |
+
|
40 |
+
During the screening framework beyond the antibiotics/enzyme-inhibitors, the antiviral
|
41 |
+
abacavir was detected. Therefore,
|
42 |
+
|
43 |
+
given the relevance of this compound, it was included in the present study. Although
|
44 |
+
enzyme inhibitors belong to the
|
45 |
+
|
46 |
+
antibiotic group, their specific pharmacological properties and detection were
|
47 |
+
sorted apart. In the present study, antibiotic
|
48 |
+
|
49 |
+
metabolites were not detected in the water bodies.
|
50 |
+
|
51 |
+
2.1.1. Antibiotics/Enzyme-Inhibitors and
|
52 |
+
|
53 |
+
Abacavir in Surface-Water'
|
54 |
+
- 'surface water. The relatively higher detection of substances downstream of the
|
55 |
+
effluent discharge points compared with a
|
56 |
+
|
57 |
+
low detection in upstream samples could be attributed to the low efficiency in
|
58 |
+
urban wastewater treatment plants or
|
59 |
+
|
60 |
+
agricultural pressure. The environmental impact is more critical due to active
|
61 |
+
substances in drinking water or premix
|
62 |
+
|
63 |
+
medicated feeds in the veterinary site.
|
64 |
+
|
65 |
+
Furthermore, the detection of substances of exclusive human use (abacavir, tazobactam
|
66 |
+
and cilastatin) prove the weak'
|
67 |
+
- source_sentence: What group of pharmaceuticals was sulfamethazine matched to when
|
68 |
+
its quantity was missing?
|
69 |
+
sentences:
|
70 |
+
- 'ciprofloxacin
|
71 |
+
|
72 |
+
43%
|
73 |
+
|
74 |
+
(3/7), enrofloxacin, norfloxacin, trimethoprim, lincomycin (29% (2/7), abacavir
|
75 |
+
and tetracycline
|
76 |
+
|
77 |
+
14% (1/7). The enzyme inhibitors, namely clavulanic acid and cilastatin, were
|
78 |
+
detected once in an urban region located
|
79 |
+
|
80 |
+
well. This catchment point showed the most significant
|
81 |
+
|
82 |
+
number of pharmaceuticals. West/Tejo and Centre were the regions with the most
|
83 |
+
considerable number of substances in
|
84 |
+
|
85 |
+
groundwater, accounting for 43%. All groundwater
|
86 |
+
|
87 |
+
samples were contaminated by at least one antibiotic. Supplemental Tables S2 and
|
88 |
+
S4 contain a detailed description of
|
89 |
+
|
90 |
+
the'
|
91 |
+
- 'clarithromycin) were the only ones that demonstrated the potential to concentrate
|
92 |
+
in living organisms (log Kow ≥ 3) [14].
|
93 |
+
|
94 |
+
All the remaining antibiotics showed a relatively low log Kow and were expected
|
95 |
+
to be present mainly in surface water.
|
96 |
+
|
97 |
+
However, the soil mobility/adsorption detected The detected pharmaceuticals showed
|
98 |
+
high to moderate water solubility
|
99 |
+
|
100 |
+
and are small ionisable molecules (MW ≤ 900 g/mol). Regarding the octanol/water
|
101 |
+
partitioning coefficient (log Kow) data,'
|
102 |
+
- 'missing quantity for sulfamethazine, the sulfonamides group has been matched.
|
103 |
+
|
104 |
+
Consumption (Kg) of the detected pharmaceuticals in Portugal (2017).
|
105 |
+
|
106 |
+
1 Amount from ESVAC Report-2017; 2 Match the sulfonamides amount; NA-not available.
|
107 |
+
|
108 |
+
Amount of detected pharmaceuticals consumption per Portuguese region. Amount of
|
109 |
+
detected pharmaceuticals
|
110 |
+
|
111 |
+
consumption per Portuguese region.'
|
112 |
+
- source_sentence: What directive sets environmental quality standards for substances
|
113 |
+
in surface waters?
|
114 |
+
sentences:
|
115 |
+
- 'As much as the specificities of each member state should be considered this issue
|
116 |
+
has become one of the European
|
117 |
+
|
118 |
+
community''s main concerns [8].
|
119 |
+
|
120 |
+
The strategies against water pollution are provided in the Water Framework Directive
|
121 |
+
[9] and the Directive on
|
122 |
+
|
123 |
+
Environmental Quality Standards that set environmental quality standards (EQS)
|
124 |
+
for the substances in surface waters
|
125 |
+
|
126 |
+
and confirm their designation as priority or priority hazardous substances [10].
|
127 |
+
Evidence of potential impacts and'
|
128 |
+
- 'seems to undertake a similar fate in the environment.
|
129 |
+
|
130 |
+
Nevertheless, due to stronger adsorption, with higher emergence in sediment, its
|
131 |
+
occurrence in the surface water is lower
|
132 |
+
|
133 |
+
[71]. The use of tetracyclines, mainly as medicated premix and oral solution for
|
134 |
+
food-producing animals [72], and the very
|
135 |
+
|
136 |
+
low bioavailability (e.g. in pig feed) [43] contribute to increasing its release
|
137 |
+
into the environment. Regarding macrolides,
|
138 |
+
|
139 |
+
erythromycin and clarithromycin exhibit a remarkable frequency of detection in
|
140 |
+
surface water samples. The most'
|
141 |
+
- 'low flows; otherwise, POCIS might be damage. In ground-waters was used one POCIS
|
142 |
+
unit/well. Due to the high sorption
|
143 |
+
|
144 |
+
capacity, POCIS was deployed approximately for 30 days, allowing the polar organic
|
145 |
+
compounds adsorbed to be in the
|
146 |
+
|
147 |
+
equilibrium stage with the active substances in an aqueous medium. In the laboratory,
|
148 |
+
POCIS disks were frozen until
|
149 |
+
|
150 |
+
extraction.
|
151 |
+
|
152 |
+
4.2.2. Qualitative Analysis Method Used
|
153 |
+
|
154 |
+
for the Characterisation of Antibiotics in
|
155 |
+
|
156 |
+
Surface-Groundwater'
|
157 |
+
- source_sentence: What is the molecular weight range of the detected pharmaceuticals?
|
158 |
+
sentences:
|
159 |
+
- '2.3. Physicochemical Properties and Key Pharmacokinetic Features of Detected
|
160 |
+
Pharmaceuticals 2.3. Physicochemical
|
161 |
+
|
162 |
+
Properties and Key Pharmacokinetic Features of Detected Pharmaceuticals
|
163 |
+
|
164 |
+
The detected pharmaceuticals showed high to moderate water solubility and are
|
165 |
+
small ionisable molecules (MW ≤ 900
|
166 |
+
|
167 |
+
g/mol). Regarding the octanol/water partitioning coefficient (log Kow) data, macrolide
|
168 |
+
antibiotics (azithromycin and
|
169 |
+
|
170 |
+
clarithromycin) were the only ones that demonstrated the potential to concentrate
|
171 |
+
in living organisms (log Kow ≥ 3) [14].'
|
172 |
+
- 'As much as the specificities of each member state should be considered this issue
|
173 |
+
has become one of the European
|
174 |
+
|
175 |
+
community''s main concerns [8].
|
176 |
+
|
177 |
+
The strategies against water pollution are provided in the Water Framework Directive
|
178 |
+
[9] and the Directive on
|
179 |
+
|
180 |
+
Environmental Quality Standards that set environmental quality standards (EQS)
|
181 |
+
for the substances in surface waters
|
182 |
+
|
183 |
+
and confirm their designation as priority or priority hazardous substances [10].
|
184 |
+
Evidence of potential impacts and'
|
185 |
+
- 'passive samplers in groundwater considered the well technical features; the depth
|
186 |
+
and groundwater level were previously
|
187 |
+
|
188 |
+
determined since they should be detected at the superficial levels. The passive
|
189 |
+
sampler was placed using a water level
|
190 |
+
|
191 |
+
meter, 2 m below the groundwater level. The sampler always remained immersed in
|
192 |
+
water, avoiding extractions and the
|
193 |
+
|
194 |
+
regional lowering of the water table [104]. For the sampling stations, sites of
|
195 |
+
different environmental pressures were
|
196 |
+
|
197 |
+
considered, specifically urban, agricultural area/animal production, and aquaculture.
|
198 |
+
The information regarding the'
|
199 |
+
- source_sentence: What was the most frequently identified pharmaceutical in the groundwater
|
200 |
+
samples?
|
201 |
+
sentences:
|
202 |
+
- 'Pharmacokinetic characteristics may represent key features in understanding antibiotics
|
203 |
+
occurrence [62]. Most antibiotics
|
204 |
+
|
205 |
+
are not completely metabolised in humans and animals; thus, a high percentage
|
206 |
+
of the active substance (40-90%) is
|
207 |
+
|
208 |
+
excreted in urine/faeces in the unchanged form. These molecules are discharged
|
209 |
+
into water and soil through wastewater,
|
210 |
+
|
211 |
+
animal manure, and sewage sludge, frequently used as fertilisers to agricultural
|
212 |
+
lands. Also, it is expected that the
|
213 |
+
|
214 |
+
hospital effluent will contribute partly to the pharmaceutical load in the wastewater
|
215 |
+
treatment plant influence [63].'
|
216 |
+
- 'many domestic and livestock animals. Several formulations of powder for administration
|
217 |
+
in drinking water and medicated
|
218 |
+
|
219 |
+
premix are available for poultry and pigs. The excretion of amoxicillin is predominantly
|
220 |
+
renal; more than 80% of the parent
|
221 |
+
|
222 |
+
drug is recovered unchanged in the urine. While bioavailability of 75 to 80% is
|
223 |
+
reported in humans, a low value (~30%)
|
224 |
+
|
225 |
+
was observed in pigs, calves, foals, and pigeons [26,52]. Maybe this last group
|
226 |
+
of animals contribute more sharply to the'
|
227 |
+
- 'from one to five compounds. The most frequently identified pharmaceuticals, in
|
228 |
+
decreasing order, were ciprofloxacin 43%
|
229 |
+
|
230 |
+
(3/7), enrofloxacin, norfloxacin, trimethoprim, lincomycin (29% (2/7), abacavir
|
231 |
+
and tetracycline 14% (1/7). The enzyme
|
232 |
+
|
233 |
+
inhibitors, namely clavulanic acid and cilastatin, were detected once in an urban
|
234 |
+
region located well. This catchment point
|
235 |
+
|
236 |
+
showed the most significant number of pharmaceuticals. West/Tejo and Centre were
|
237 |
+
the regions with the most
|
238 |
+
|
239 |
+
considerable number of substances in groundwater, accounting for 43%. All groundwater
|
240 |
+
samples were contaminated by'
|
241 |
+
pipeline_tag: sentence-similarity
|
242 |
+
library_name: sentence-transformers
|
243 |
+
metrics:
|
244 |
+
- cosine_accuracy
|
245 |
+
model-index:
|
246 |
+
- name: SentenceTransformer based on microsoft/mpnet-base
|
247 |
+
results:
|
248 |
+
- task:
|
249 |
+
type: triplet
|
250 |
+
name: Triplet
|
251 |
+
dataset:
|
252 |
+
name: initial test
|
253 |
+
type: initial_test
|
254 |
+
metrics:
|
255 |
+
- type: cosine_accuracy
|
256 |
+
value: 0.9599999785423279
|
257 |
+
name: Cosine Accuracy
|
258 |
+
- task:
|
259 |
+
type: triplet
|
260 |
+
name: Triplet
|
261 |
+
dataset:
|
262 |
+
name: final test
|
263 |
+
type: final_test
|
264 |
+
metrics:
|
265 |
+
- type: cosine_accuracy
|
266 |
+
value: 0.6800000071525574
|
267 |
+
name: Cosine Accuracy
|
268 |
+
- type: cosine_accuracy
|
269 |
+
value: 0.8999999761581421
|
270 |
+
name: Cosine Accuracy
|
271 |
+
- type: cosine_accuracy
|
272 |
+
value: 0.9200000166893005
|
273 |
+
name: Cosine Accuracy
|
274 |
+
- type: cosine_accuracy
|
275 |
+
value: 0.9399999976158142
|
276 |
+
name: Cosine Accuracy
|
277 |
+
- type: cosine_accuracy
|
278 |
+
value: 0.9599999785423279
|
279 |
+
name: Cosine Accuracy
|
280 |
+
- type: cosine_accuracy
|
281 |
+
value: 0.9599999785423279
|
282 |
+
name: Cosine Accuracy
|
283 |
+
- type: cosine_accuracy
|
284 |
+
value: 0.9599999785423279
|
285 |
+
name: Cosine Accuracy
|
286 |
+
- type: cosine_accuracy
|
287 |
+
value: 0.9599999785423279
|
288 |
+
name: Cosine Accuracy
|
289 |
+
- type: cosine_accuracy
|
290 |
+
value: 0.9599999785423279
|
291 |
+
name: Cosine Accuracy
|
292 |
+
- type: cosine_accuracy
|
293 |
+
value: 0.9800000190734863
|
294 |
+
name: Cosine Accuracy
|
295 |
+
- type: cosine_accuracy
|
296 |
+
value: 0.9800000190734863
|
297 |
+
name: Cosine Accuracy
|
298 |
+
---
|
299 |
+
|
300 |
+
# SentenceTransformer based on microsoft/mpnet-base
|
301 |
+
|
302 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base) on the json dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
|
303 |
+
|
304 |
+
## Model Details
|
305 |
+
|
306 |
+
### Model Description
|
307 |
+
- **Model Type:** Sentence Transformer
|
308 |
+
- **Base model:** [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base) <!-- at revision 6996ce1e91bd2a9c7d7f61daec37463394f73f09 -->
|
309 |
+
- **Maximum Sequence Length:** 512 tokens
|
310 |
+
- **Output Dimensionality:** 768 dimensions
|
311 |
+
- **Similarity Function:** Cosine Similarity
|
312 |
+
- **Training Dataset:**
|
313 |
+
- json
|
314 |
+
<!-- - **Language:** Unknown -->
|
315 |
+
<!-- - **License:** Unknown -->
|
316 |
+
|
317 |
+
### Model Sources
|
318 |
+
|
319 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
320 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
321 |
+
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
|
322 |
+
|
323 |
+
### Full Model Architecture
|
324 |
+
|
325 |
+
```
|
326 |
+
SentenceTransformer(
|
327 |
+
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False, 'architecture': 'MPNetModel'})
|
328 |
+
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
|
329 |
+
)
|
330 |
+
```
|
331 |
+
|
332 |
+
## Usage
|
333 |
+
|
334 |
+
### Direct Usage (Sentence Transformers)
|
335 |
+
|
336 |
+
First install the Sentence Transformers library:
|
337 |
+
|
338 |
+
```bash
|
339 |
+
pip install -U sentence-transformers
|
340 |
+
```
|
341 |
+
|
342 |
+
Then you can load this model and run inference.
|
343 |
+
```python
|
344 |
+
from sentence_transformers import SentenceTransformer
|
345 |
+
|
346 |
+
# Download from the 🤗 Hub
|
347 |
+
model = SentenceTransformer("sahithkumar7/mpnet-base-finetuned-iter01")
|
348 |
+
# Run inference
|
349 |
+
sentences = [
|
350 |
+
'What was the most frequently identified pharmaceutical in the groundwater samples?',
|
351 |
+
'from one to five compounds. The most frequently identified pharmaceuticals, in decreasing order, were ciprofloxacin 43%\n(3/7), enrofloxacin, norfloxacin, trimethoprim, lincomycin (29% (2/7), abacavir and tetracycline 14% (1/7). The enzyme\ninhibitors, namely clavulanic acid and cilastatin, were detected once in an urban region located well. This catchment point\nshowed the most significant number of pharmaceuticals. West/Tejo and Centre were the regions with the most\nconsiderable number of substances in groundwater, accounting for 43%. All groundwater samples were contaminated by',
|
352 |
+
'Pharmacokinetic characteristics may represent key features in understanding antibiotics occurrence [62]. Most antibiotics\nare not completely metabolised in humans and animals; thus, a high percentage of the active substance (40-90%) is\nexcreted in urine/faeces in the unchanged form. These molecules are discharged into water and soil through wastewater,\nanimal manure, and sewage sludge, frequently used as fertilisers to agricultural lands. Also, it is expected that the\nhospital effluent will contribute partly to the pharmaceutical load in the wastewater treatment plant influence [63].',
|
353 |
+
]
|
354 |
+
embeddings = model.encode(sentences)
|
355 |
+
print(embeddings.shape)
|
356 |
+
# [3, 768]
|
357 |
+
|
358 |
+
# Get the similarity scores for the embeddings
|
359 |
+
similarities = model.similarity(embeddings, embeddings)
|
360 |
+
print(similarities)
|
361 |
+
# tensor([[ 1.0000, 0.4988, -0.0391],
|
362 |
+
# [ 0.4988, 1.0000, 0.0047],
|
363 |
+
# [-0.0391, 0.0047, 1.0000]])
|
364 |
+
```
|
365 |
+
|
366 |
+
<!--
|
367 |
+
### Direct Usage (Transformers)
|
368 |
+
|
369 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
370 |
+
|
371 |
+
</details>
|
372 |
+
-->
|
373 |
+
|
374 |
+
<!--
|
375 |
+
### Downstream Usage (Sentence Transformers)
|
376 |
+
|
377 |
+
You can finetune this model on your own dataset.
|
378 |
+
|
379 |
+
<details><summary>Click to expand</summary>
|
380 |
+
|
381 |
+
</details>
|
382 |
+
-->
|
383 |
+
|
384 |
+
<!--
|
385 |
+
### Out-of-Scope Use
|
386 |
+
|
387 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
388 |
+
-->
|
389 |
+
|
390 |
+
## Evaluation
|
391 |
+
|
392 |
+
### Metrics
|
393 |
+
|
394 |
+
#### Triplet
|
395 |
+
|
396 |
+
* Datasets: `initial_test`, `final_test`, `final_test`, `final_test`, `final_test`, `final_test`, `final_test`, `final_test`, `final_test`, `final_test`, `final_test` and `final_test`
|
397 |
+
* Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
|
398 |
+
|
399 |
+
| Metric | initial_test | final_test |
|
400 |
+
|:--------------------|:-------------|:-----------|
|
401 |
+
| **cosine_accuracy** | **0.96** | **0.98** |
|
402 |
+
|
403 |
+
<!--
|
404 |
+
## Bias, Risks and Limitations
|
405 |
+
|
406 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
407 |
+
-->
|
408 |
+
|
409 |
+
<!--
|
410 |
+
### Recommendations
|
411 |
+
|
412 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
413 |
+
-->
|
414 |
+
|
415 |
+
## Training Details
|
416 |
+
|
417 |
+
### Training Dataset
|
418 |
+
|
419 |
+
#### json
|
420 |
+
|
421 |
+
* Dataset: json
|
422 |
+
* Size: 80 training samples
|
423 |
+
* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
|
424 |
+
* Approximate statistics based on the first 80 samples:
|
425 |
+
| | anchor | positive | negative |
|
426 |
+
|:--------|:----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
|
427 |
+
| type | string | string | string |
|
428 |
+
| details | <ul><li>min: 9 tokens</li><li>mean: 16.14 tokens</li><li>max: 33 tokens</li></ul> | <ul><li>min: 48 tokens</li><li>mean: 125.65 tokens</li><li>max: 218 tokens</li></ul> | <ul><li>min: 48 tokens</li><li>mean: 122.97 tokens</li><li>max: 211 tokens</li></ul> |
|
429 |
+
* Samples:
|
430 |
+
| anchor | positive | negative |
|
431 |
+
|:-----------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
432 |
+
| <code>Which two macrolide antibiotics are frequently detected in surface water samples?</code> | <code>seems to undertake a similar fate in the environment.<br>Nevertheless, due to stronger adsorption, with higher emergence in sediment, its occurrence in the surface water is lower<br>[71]. The use of tetracyclines, mainly as medicated premix and oral solution for food-producing animals [72], and the very<br>low bioavailability (e.g. in pig feed) [43] contribute to increasing its release into the environment. Regarding macrolides,<br>erythromycin and clarithromycin exhibit a remarkable frequency of detection in surface water samples. The most</code> | <code>Nonetheless, besides the sorption capacity, these antibiotics have high solubility in water. Crucial routes for these<br>substances into the environment are manure from animal production and sewage sludge from wastewater treatment<br>plant (WWTP) used as fertilisers. Therefore, these substances have been evidenced in topsoil samples [68]. These<br>quinolones and other antibiotics, for instance, norfloxacin and tetracycline, have been identified in groundwater samples<br>despite being influenced by sorption processes. They were not readily degraded; instead, the input into groundwater</code> |
|
433 |
+
| <code>What antimicrobial drugs were identified in the survey besides macrolides?</code> | <code>is one of the most frequently pharmaceutical in representative rivers [74,75]. The three macrolides identified in our<br>detection survey are included since 2018 in the first 'watch list' [76].<br>Another group of antimicrobial drugs identified in our survey were sulfamethoxazole/trimethoprim and sulfamethazine.<br>Sulfamethoxazole/trimethoprim are often used combined since the effectiveness of sulfonamides is enhanced. In the<br>present study, the detection of both substances was comparable; however, trimethoprim was detected in groundwater.</code> | <code>upstream samples obtained in rural locations was demonstrated and could be attributed to a low efficiency in the urban<br>wastewater treatment plants or due to agricultural pressure.<br>The higher frequency of detection for most substances was observed in the Ave river and Ria Formosa, confirming that<br>several effluents impact these water bodies from urban wastewater treatment plants and livestock production.<br>Pharmacokinetic characteristics may represent key features in understanding antibiotics occurrence [62]. Most antibiotics</code> |
|
434 |
+
| <code>How long was the observational period of the antibiotic survey in Portugal?</code> | <code>of antibiotics and their metabolites in surface- groundwater. It seeks to reflect the current demographic, spatial, drug<br>consumption, and drug profile on an observational period of 3 years in Portugal. The greatest challenge of this survey<br>data will be to promote the ecopharmacovigilance framework development shortly to implement measures for avoiding<br>misuse/overuse of antibiotics and slow down emission and antibiotic resistance.<br>2. Results<br>2.1. Frequency of Detections:<br>Antibiotics/Enzyme-Inhibitors and Abacavir<br>in Surface-Groundwater</code> | <code>despite being influenced by sorption processes. They were not readily degraded; instead, the input into groundwater<br>could be due to livestock farming pressure, namely by spreading manure in the soil or the possible sewage sludge<br>application in the area. High clay and low sand content in soils can decrease the mobility of pharmaceuticals, which is<br>attributed to clay intense exchange capacity. Thus, soil properties (e.g. particle composition) are a significant, influential</code> |
|
435 |
+
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
|
436 |
+
```json
|
437 |
+
{
|
438 |
+
"scale": 20.0,
|
439 |
+
"similarity_fct": "cos_sim"
|
440 |
+
}
|
441 |
+
```
|
442 |
+
|
443 |
+
### Evaluation Dataset
|
444 |
+
|
445 |
+
#### json
|
446 |
+
|
447 |
+
* Dataset: json
|
448 |
+
* Size: 20 evaluation samples
|
449 |
+
* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
|
450 |
+
* Approximate statistics based on the first 20 samples:
|
451 |
+
| | anchor | positive | negative |
|
452 |
+
|:--------|:----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
|
453 |
+
| type | string | string | string |
|
454 |
+
| details | <ul><li>min: 11 tokens</li><li>mean: 16.4 tokens</li><li>max: 25 tokens</li></ul> | <ul><li>min: 76 tokens</li><li>mean: 113.65 tokens</li><li>max: 148 tokens</li></ul> | <ul><li>min: 89 tokens</li><li>mean: 118.8 tokens</li><li>max: 162 tokens</li></ul> |
|
455 |
+
* Samples:
|
456 |
+
| anchor | positive | negative |
|
457 |
+
|:-----------------------------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
458 |
+
| <code>What percentage of unchanged excretion did the most significant number of detected substances show?</code> | <code>coefficients were not available for lincomycin, clavulanic acid and cilastatin.<br>Physicochemical properties of detected pharmaceuticals.<br>1 Data retrieved from [16]; 2 Data retrieved from [17]; 3 Data retrieved from [18]; 4 Data retrieved from [19]; 5<br>Data retrieved from [20];<br>6 Data retrieved from [21]; 7 Data retrieved from [22]; 8 Data retrieved from [23]; 9 Data retrieved from [24]; 10<br>Data retrieved from [25];<br>NA-not available.<br>The most significant number of detected substances showed a percentage of unchanged excretion higher than 40%.</code> | <code>1. Introduction<br>Antibiotics are a critical component of human and veterinary modern medicine, developed to produce desirable or<br>beneficial effects on infections induced by pathogens. Like most pharmaceuticals, antibiotics tend to be small organic<br>polar compounds, generally ionisable, ordinarily subject to a metabolism or biotransformation process by the organism to<br>be eliminated more efficiently [1,2]. The excretion of these compounds and their metabolites occurs mainly through urine,</code> |
|
459 |
+
| <code>How many kilograms of abacavir were detected in Portugal in 2017?</code> | <code>Regarding the different regions, it has been concluded that North and West/Tejo were the regions with the higher<br>consuming values. Both regions presented a significant value (33%) for the abacavir. For the detected antiviral abacavir,<br>an amount of 1458 kg has been observed.<br>Regarding antibiotics used in veterinary medicine, the regional amount was not available. Likewise, due to the reported<br>missing quantity for sulfamethazine, the sulfonamides group has been matched.<br>Consumption (Kg) of the detected pharmaceuticals in Portugal (2017).</code> | <code>43%<br>(3/7), enrofloxacin, norfloxacin, trimethoprim, lincomycin (29% (2/7), abacavir and tetracycline<br>14% (1/7). The enzyme inhibitors, namely clavulanic acid and cilastatin, were detected once in an urban region located<br>well. This catchment point showed the most significant<br>number of pharmaceuticals. West/Tejo and Centre were the regions with the most considerable number of substances in<br>groundwater, accounting for 43%. All groundwater<br>samples were contaminated by at least one antibiotic. Supplemental Tables S2 and S4 contain a detailed description of<br>the</code> |
|
460 |
+
| <code>What must marketing authorisation procedures for medicines include since 2006?</code> | <code>substances in passive samplers [7]. Since 2006, marketing authorisation procedures for both human and veterinary<br>medicines must include an environmental risk assessment that comprises a prospective exposure assessment,<br>underestimating the possible impact and the occurrence of antibiotics after years of consumption. Ultimately, the potential<br>risk may not be correctly anticipated. It becomes urgent to generate new data, mainly to refine exposure assessments.<br>As much as the specificities of each member state should be considered this issue has become one of the European</code> | <code>clarithromycin/erythromycin, tetracycline, sulfamethoxazole, and abacavir. In groundwater, enrofloxacin/ciprofloxacin,<br>norfloxacin, trimethoprim, lincomycin, abacavir and tetracycline were recovered. Metabolites were not detected in water<br>bodies. Noticeable was the detection of enzyme inhibitors, tazobactam and cilastatin, which are both for exclusive<br>hospital use. The North region and Algarve (South) were the areas with the most significant frequency of substances in<br>surface water. The relatively higher detection of substances downstream of the effluent discharge points compared with a</code> |
|
461 |
+
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
|
462 |
+
```json
|
463 |
+
{
|
464 |
+
"scale": 20.0,
|
465 |
+
"similarity_fct": "cos_sim"
|
466 |
+
}
|
467 |
+
```
|
468 |
+
|
469 |
+
### Training Hyperparameters
|
470 |
+
#### Non-Default Hyperparameters
|
471 |
+
|
472 |
+
- `eval_strategy`: steps
|
473 |
+
- `per_device_train_batch_size`: 16
|
474 |
+
- `per_device_eval_batch_size`: 16
|
475 |
+
- `num_train_epochs`: 1
|
476 |
+
- `warmup_ratio`: 0.1
|
477 |
+
- `fp16`: True
|
478 |
+
- `batch_sampler`: no_duplicates
|
479 |
+
|
480 |
+
#### All Hyperparameters
|
481 |
+
<details><summary>Click to expand</summary>
|
482 |
+
|
483 |
+
- `overwrite_output_dir`: False
|
484 |
+
- `do_predict`: False
|
485 |
+
- `eval_strategy`: steps
|
486 |
+
- `prediction_loss_only`: True
|
487 |
+
- `per_device_train_batch_size`: 16
|
488 |
+
- `per_device_eval_batch_size`: 16
|
489 |
+
- `per_gpu_train_batch_size`: None
|
490 |
+
- `per_gpu_eval_batch_size`: None
|
491 |
+
- `gradient_accumulation_steps`: 1
|
492 |
+
- `eval_accumulation_steps`: None
|
493 |
+
- `torch_empty_cache_steps`: None
|
494 |
+
- `learning_rate`: 5e-05
|
495 |
+
- `weight_decay`: 0.0
|
496 |
+
- `adam_beta1`: 0.9
|
497 |
+
- `adam_beta2`: 0.999
|
498 |
+
- `adam_epsilon`: 1e-08
|
499 |
+
- `max_grad_norm`: 1.0
|
500 |
+
- `num_train_epochs`: 1
|
501 |
+
- `max_steps`: -1
|
502 |
+
- `lr_scheduler_type`: linear
|
503 |
+
- `lr_scheduler_kwargs`: {}
|
504 |
+
- `warmup_ratio`: 0.1
|
505 |
+
- `warmup_steps`: 0
|
506 |
+
- `log_level`: passive
|
507 |
+
- `log_level_replica`: warning
|
508 |
+
- `log_on_each_node`: True
|
509 |
+
- `logging_nan_inf_filter`: True
|
510 |
+
- `save_safetensors`: True
|
511 |
+
- `save_on_each_node`: False
|
512 |
+
- `save_only_model`: False
|
513 |
+
- `restore_callback_states_from_checkpoint`: False
|
514 |
+
- `no_cuda`: False
|
515 |
+
- `use_cpu`: False
|
516 |
+
- `use_mps_device`: False
|
517 |
+
- `seed`: 42
|
518 |
+
- `data_seed`: None
|
519 |
+
- `jit_mode_eval`: False
|
520 |
+
- `use_ipex`: False
|
521 |
+
- `bf16`: False
|
522 |
+
- `fp16`: True
|
523 |
+
- `fp16_opt_level`: O1
|
524 |
+
- `half_precision_backend`: auto
|
525 |
+
- `bf16_full_eval`: False
|
526 |
+
- `fp16_full_eval`: False
|
527 |
+
- `tf32`: None
|
528 |
+
- `local_rank`: 0
|
529 |
+
- `ddp_backend`: None
|
530 |
+
- `tpu_num_cores`: None
|
531 |
+
- `tpu_metrics_debug`: False
|
532 |
+
- `debug`: []
|
533 |
+
- `dataloader_drop_last`: False
|
534 |
+
- `dataloader_num_workers`: 0
|
535 |
+
- `dataloader_prefetch_factor`: None
|
536 |
+
- `past_index`: -1
|
537 |
+
- `disable_tqdm`: False
|
538 |
+
- `remove_unused_columns`: True
|
539 |
+
- `label_names`: None
|
540 |
+
- `load_best_model_at_end`: False
|
541 |
+
- `ignore_data_skip`: False
|
542 |
+
- `fsdp`: []
|
543 |
+
- `fsdp_min_num_params`: 0
|
544 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
545 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
546 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
|
547 |
+
- `deepspeed`: None
|
548 |
+
- `label_smoothing_factor`: 0.0
|
549 |
+
- `optim`: adamw_torch
|
550 |
+
- `optim_args`: None
|
551 |
+
- `adafactor`: False
|
552 |
+
- `group_by_length`: False
|
553 |
+
- `length_column_name`: length
|
554 |
+
- `ddp_find_unused_parameters`: None
|
555 |
+
- `ddp_bucket_cap_mb`: None
|
556 |
+
- `ddp_broadcast_buffers`: False
|
557 |
+
- `dataloader_pin_memory`: True
|
558 |
+
- `dataloader_persistent_workers`: False
|
559 |
+
- `skip_memory_metrics`: True
|
560 |
+
- `use_legacy_prediction_loop`: False
|
561 |
+
- `push_to_hub`: False
|
562 |
+
- `resume_from_checkpoint`: None
|
563 |
+
- `hub_model_id`: None
|
564 |
+
- `hub_strategy`: every_save
|
565 |
+
- `hub_private_repo`: None
|
566 |
+
- `hub_always_push`: False
|
567 |
+
- `gradient_checkpointing`: False
|
568 |
+
- `gradient_checkpointing_kwargs`: None
|
569 |
+
- `include_inputs_for_metrics`: False
|
570 |
+
- `include_for_metrics`: []
|
571 |
+
- `eval_do_concat_batches`: True
|
572 |
+
- `fp16_backend`: auto
|
573 |
+
- `push_to_hub_model_id`: None
|
574 |
+
- `push_to_hub_organization`: None
|
575 |
+
- `mp_parameters`:
|
576 |
+
- `auto_find_batch_size`: False
|
577 |
+
- `full_determinism`: False
|
578 |
+
- `torchdynamo`: None
|
579 |
+
- `ray_scope`: last
|
580 |
+
- `ddp_timeout`: 1800
|
581 |
+
- `torch_compile`: False
|
582 |
+
- `torch_compile_backend`: None
|
583 |
+
- `torch_compile_mode`: None
|
584 |
+
- `include_tokens_per_second`: False
|
585 |
+
- `include_num_input_tokens_seen`: False
|
586 |
+
- `neftune_noise_alpha`: None
|
587 |
+
- `optim_target_modules`: None
|
588 |
+
- `batch_eval_metrics`: False
|
589 |
+
- `eval_on_start`: False
|
590 |
+
- `use_liger_kernel`: False
|
591 |
+
- `eval_use_gather_object`: False
|
592 |
+
- `average_tokens_across_devices`: False
|
593 |
+
- `prompts`: None
|
594 |
+
- `batch_sampler`: no_duplicates
|
595 |
+
- `multi_dataset_batch_sampler`: proportional
|
596 |
+
- `router_mapping`: {}
|
597 |
+
- `learning_rate_mapping`: {}
|
598 |
+
|
599 |
+
</details>
|
600 |
+
|
601 |
+
### Training Logs
|
602 |
+
<details><summary>Click to expand</summary>
|
603 |
+
|
604 |
+
| Epoch | Step | Training Loss | Validation Loss | initial_test_cosine_accuracy | final_test_cosine_accuracy |
|
605 |
+
|:-----:|:----:|:-------------:|:---------------:|:----------------------------:|:--------------------------:|
|
606 |
+
| -1 | -1 | - | - | 0.7800 | - |
|
607 |
+
| 0.2 | 1 | 3.3315 | - | - | - |
|
608 |
+
| 0.4 | 2 | 3.0922 | - | - | - |
|
609 |
+
| 0.6 | 3 | 3.2635 | - | - | - |
|
610 |
+
| 0.8 | 4 | 3.0702 | - | - | - |
|
611 |
+
| 1.0 | 5 | 3.3282 | - | - | - |
|
612 |
+
| -1 | -1 | - | - | - | 0.6800 |
|
613 |
+
| 0.2 | 1 | 2.9487 | - | - | - |
|
614 |
+
| 0.4 | 2 | 2.9845 | - | - | - |
|
615 |
+
| 0.6 | 3 | 2.935 | - | - | - |
|
616 |
+
| 0.8 | 4 | 3.0702 | - | - | - |
|
617 |
+
| 1.0 | 5 | 3.0039 | - | - | - |
|
618 |
+
| 1.2 | 6 | 2.4806 | - | - | - |
|
619 |
+
| 1.4 | 7 | 2.2646 | - | - | - |
|
620 |
+
| 1.6 | 8 | 1.8101 | - | - | - |
|
621 |
+
| 1.8 | 9 | 1.3463 | - | - | - |
|
622 |
+
| 2.0 | 10 | 1.942 | - | - | - |
|
623 |
+
| -1 | -1 | - | - | - | 0.9000 |
|
624 |
+
| 0.2 | 1 | 2.2356 | - | - | - |
|
625 |
+
| 0.4 | 2 | 1.0123 | - | - | - |
|
626 |
+
| 0.6 | 3 | 1.2411 | - | - | - |
|
627 |
+
| 0.8 | 4 | 0.9194 | - | - | - |
|
628 |
+
| 1.0 | 5 | 0.891 | - | - | - |
|
629 |
+
| 1.2 | 6 | 0.602 | - | - | - |
|
630 |
+
| 1.4 | 7 | 0.5426 | - | - | - |
|
631 |
+
| 1.6 | 8 | 0.5738 | - | - | - |
|
632 |
+
| 1.8 | 9 | 0.2678 | - | - | - |
|
633 |
+
| 2.0 | 10 | 0.7113 | - | - | - |
|
634 |
+
| 2.2 | 11 | 0.2911 | - | - | - |
|
635 |
+
| 2.4 | 12 | 0.4745 | - | - | - |
|
636 |
+
| 2.6 | 13 | 0.4188 | - | - | - |
|
637 |
+
| 2.8 | 14 | 0.3708 | - | - | - |
|
638 |
+
| 3.0 | 15 | 0.2882 | - | - | - |
|
639 |
+
| -1 | -1 | - | - | - | 0.9200 |
|
640 |
+
| 0.2 | 1 | 0.5156 | - | - | - |
|
641 |
+
| 0.4 | 2 | 0.0749 | - | - | - |
|
642 |
+
| 0.6 | 3 | 0.0634 | - | - | - |
|
643 |
+
| 0.8 | 4 | 0.0534 | - | - | - |
|
644 |
+
| 1.0 | 5 | 0.019 | - | - | - |
|
645 |
+
| 1.2 | 6 | 0.0682 | - | - | - |
|
646 |
+
| 1.4 | 7 | 0.0381 | - | - | - |
|
647 |
+
| 1.6 | 8 | 0.171 | - | - | - |
|
648 |
+
| 1.8 | 9 | 0.1188 | - | - | - |
|
649 |
+
| 2.0 | 10 | 0.1861 | - | - | - |
|
650 |
+
| 2.2 | 11 | 0.0895 | - | - | - |
|
651 |
+
| 2.4 | 12 | 0.2492 | - | - | - |
|
652 |
+
| 2.6 | 13 | 0.0964 | - | - | - |
|
653 |
+
| 2.8 | 14 | 0.2424 | - | - | - |
|
654 |
+
| 3.0 | 15 | 0.1096 | - | - | - |
|
655 |
+
| 3.2 | 16 | 0.1981 | - | - | - |
|
656 |
+
| 3.4 | 17 | 0.1438 | - | - | - |
|
657 |
+
| 3.6 | 18 | 0.3454 | - | - | - |
|
658 |
+
| 3.8 | 19 | 0.4011 | - | - | - |
|
659 |
+
| 4.0 | 20 | 0.1591 | 0.5567 | 0.9400 | - |
|
660 |
+
| -1 | -1 | - | - | - | 0.9400 |
|
661 |
+
| 0.125 | 1 | 0.0594 | - | - | - |
|
662 |
+
| 0.25 | 2 | 0.0584 | - | - | - |
|
663 |
+
| 0.375 | 3 | 0.0146 | - | - | - |
|
664 |
+
| 0.5 | 4 | 0.0542 | - | - | - |
|
665 |
+
| 0.625 | 5 | 0.0965 | - | - | - |
|
666 |
+
| 0.75 | 6 | 0.2209 | - | - | - |
|
667 |
+
| 0.875 | 7 | 0.0312 | - | - | - |
|
668 |
+
| 1.0 | 8 | 0.1142 | - | - | - |
|
669 |
+
| -1 | -1 | - | - | - | 0.9600 |
|
670 |
+
| 0.125 | 1 | 0.0082 | - | - | - |
|
671 |
+
| 0.25 | 2 | 0.004 | - | - | - |
|
672 |
+
| 0.375 | 3 | 0.001 | - | - | - |
|
673 |
+
| 0.5 | 4 | 0.0118 | - | - | - |
|
674 |
+
| 0.625 | 5 | 0.0508 | - | - | - |
|
675 |
+
| 0.75 | 6 | 0.0816 | - | - | - |
|
676 |
+
| 0.875 | 7 | 0.0149 | - | - | - |
|
677 |
+
| 1.0 | 8 | 0.0163 | - | - | - |
|
678 |
+
| 1.125 | 9 | 0.038 | - | - | - |
|
679 |
+
| 1.25 | 10 | 0.0618 | - | - | - |
|
680 |
+
| 1.375 | 11 | 0.0097 | - | - | - |
|
681 |
+
| 1.5 | 12 | 0.0368 | - | - | - |
|
682 |
+
| 1.625 | 13 | 0.0212 | - | - | - |
|
683 |
+
| 1.75 | 14 | 0.0072 | - | - | - |
|
684 |
+
| 1.875 | 15 | 0.0037 | - | - | - |
|
685 |
+
| 2.0 | 16 | 0.128 | - | - | - |
|
686 |
+
| -1 | -1 | - | - | - | 0.9600 |
|
687 |
+
| 0.125 | 1 | 0.0012 | - | - | - |
|
688 |
+
| 0.25 | 2 | 0.0003 | - | - | - |
|
689 |
+
| 0.375 | 3 | 0.0008 | - | - | - |
|
690 |
+
| 0.5 | 4 | 0.0008 | - | - | - |
|
691 |
+
| 0.625 | 5 | 0.0013 | - | - | - |
|
692 |
+
| 0.75 | 6 | 0.0743 | - | - | - |
|
693 |
+
| 0.875 | 7 | 0.0024 | - | - | - |
|
694 |
+
| 1.0 | 8 | 0.001 | - | - | - |
|
695 |
+
| 1.125 | 9 | 0.0024 | - | - | - |
|
696 |
+
| 1.25 | 10 | 0.01 | - | - | - |
|
697 |
+
| 1.375 | 11 | 0.0009 | - | - | - |
|
698 |
+
| 1.5 | 12 | 0.1912 | - | - | - |
|
699 |
+
| 1.625 | 13 | 0.0024 | - | - | - |
|
700 |
+
| 1.75 | 14 | 0.002 | - | - | - |
|
701 |
+
| 1.875 | 15 | 0.0038 | - | - | - |
|
702 |
+
| 2.0 | 16 | 0.1492 | - | - | - |
|
703 |
+
| 2.125 | 17 | 0.004 | - | - | - |
|
704 |
+
| 2.25 | 18 | 0.0123 | - | - | - |
|
705 |
+
| 2.375 | 19 | 0.0348 | - | - | - |
|
706 |
+
| 2.5 | 20 | 0.0068 | 0.5351 | 0.9600 | - |
|
707 |
+
| 2.625 | 21 | 0.1679 | - | - | - |
|
708 |
+
| 2.75 | 22 | 0.0123 | - | - | - |
|
709 |
+
| 2.875 | 23 | 0.0934 | - | - | - |
|
710 |
+
| 3.0 | 24 | 0.0048 | - | - | - |
|
711 |
+
| -1 | -1 | - | - | - | 0.9600 |
|
712 |
+
| 0.2 | 1 | 0.0763 | - | - | - |
|
713 |
+
| 0.4 | 2 | 0.0119 | - | - | - |
|
714 |
+
| 0.6 | 3 | 0.0019 | - | - | - |
|
715 |
+
| 0.8 | 4 | 0.0034 | - | - | - |
|
716 |
+
| 1.0 | 5 | 0.001 | - | - | - |
|
717 |
+
| -1 | -1 | - | - | - | 0.9800 |
|
718 |
+
|
719 |
+
</details>
|
720 |
+
|
721 |
+
### Framework Versions
|
722 |
+
- Python: 3.11.13
|
723 |
+
- Sentence Transformers: 5.0.0
|
724 |
+
- Transformers: 4.52.4
|
725 |
+
- PyTorch: 2.6.0+cu124
|
726 |
+
- Accelerate: 1.8.1
|
727 |
+
- Datasets: 3.6.0
|
728 |
+
- Tokenizers: 0.21.2
|
729 |
+
|
730 |
+
## Citation
|
731 |
+
|
732 |
+
### BibTeX
|
733 |
+
|
734 |
+
#### Sentence Transformers
|
735 |
+
```bibtex
|
736 |
+
@inproceedings{reimers-2019-sentence-bert,
|
737 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
738 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
739 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
740 |
+
month = "11",
|
741 |
+
year = "2019",
|
742 |
+
publisher = "Association for Computational Linguistics",
|
743 |
+
url = "https://arxiv.org/abs/1908.10084",
|
744 |
+
}
|
745 |
+
```
|
746 |
+
|
747 |
+
#### MultipleNegativesRankingLoss
|
748 |
+
```bibtex
|
749 |
+
@misc{henderson2017efficient,
|
750 |
+
title={Efficient Natural Language Response Suggestion for Smart Reply},
|
751 |
+
author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
|
752 |
+
year={2017},
|
753 |
+
eprint={1705.00652},
|
754 |
+
archivePrefix={arXiv},
|
755 |
+
primaryClass={cs.CL}
|
756 |
+
}
|
757 |
+
```
|
758 |
+
|
759 |
+
<!--
|
760 |
+
## Glossary
|
761 |
+
|
762 |
+
*Clearly define terms in order to be accessible across audiences.*
|
763 |
+
-->
|
764 |
+
|
765 |
+
<!--
|
766 |
+
## Model Card Authors
|
767 |
+
|
768 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
769 |
+
-->
|
770 |
+
|
771 |
+
<!--
|
772 |
+
## Model Card Contact
|
773 |
+
|
774 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
775 |
+
-->
|
config.json
ADDED
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"MPNetModel"
|
4 |
+
],
|
5 |
+
"attention_probs_dropout_prob": 0.1,
|
6 |
+
"bos_token_id": 0,
|
7 |
+
"eos_token_id": 2,
|
8 |
+
"hidden_act": "gelu",
|
9 |
+
"hidden_dropout_prob": 0.1,
|
10 |
+
"hidden_size": 768,
|
11 |
+
"initializer_range": 0.02,
|
12 |
+
"intermediate_size": 3072,
|
13 |
+
"layer_norm_eps": 1e-05,
|
14 |
+
"max_position_embeddings": 514,
|
15 |
+
"model_type": "mpnet",
|
16 |
+
"num_attention_heads": 12,
|
17 |
+
"num_hidden_layers": 12,
|
18 |
+
"pad_token_id": 1,
|
19 |
+
"relative_attention_num_buckets": 32,
|
20 |
+
"torch_dtype": "float32",
|
21 |
+
"transformers_version": "4.52.4",
|
22 |
+
"vocab_size": 30527
|
23 |
+
}
|
config_sentence_transformers.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"model_type": "SentenceTransformer",
|
3 |
+
"__version__": {
|
4 |
+
"sentence_transformers": "5.0.0",
|
5 |
+
"transformers": "4.52.4",
|
6 |
+
"pytorch": "2.6.0+cu124"
|
7 |
+
},
|
8 |
+
"prompts": {
|
9 |
+
"query": "",
|
10 |
+
"document": ""
|
11 |
+
},
|
12 |
+
"default_prompt_name": null,
|
13 |
+
"similarity_fn_name": "cosine"
|
14 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6a3fe86dee3b0d3265329a58a8bd973acb1baf97e46abdf20fb4da9b39ecc8b2
|
3 |
+
size 437967672
|
modules.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"idx": 0,
|
4 |
+
"name": "0",
|
5 |
+
"path": "",
|
6 |
+
"type": "sentence_transformers.models.Transformer"
|
7 |
+
},
|
8 |
+
{
|
9 |
+
"idx": 1,
|
10 |
+
"name": "1",
|
11 |
+
"path": "1_Pooling",
|
12 |
+
"type": "sentence_transformers.models.Pooling"
|
13 |
+
}
|
14 |
+
]
|
sentence_bert_config.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"max_seq_length": 512,
|
3 |
+
"do_lower_case": false
|
4 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": {
|
3 |
+
"content": "<s>",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"cls_token": {
|
10 |
+
"content": "<s>",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": true,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"eos_token": {
|
17 |
+
"content": "</s>",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"mask_token": {
|
24 |
+
"content": "<mask>",
|
25 |
+
"lstrip": true,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"pad_token": {
|
31 |
+
"content": "<pad>",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
},
|
37 |
+
"sep_token": {
|
38 |
+
"content": "</s>",
|
39 |
+
"lstrip": false,
|
40 |
+
"normalized": true,
|
41 |
+
"rstrip": false,
|
42 |
+
"single_word": false
|
43 |
+
},
|
44 |
+
"unk_token": {
|
45 |
+
"content": "[UNK]",
|
46 |
+
"lstrip": false,
|
47 |
+
"normalized": false,
|
48 |
+
"rstrip": false,
|
49 |
+
"single_word": false
|
50 |
+
}
|
51 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "<pad>",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "</s>",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "<unk>",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": true,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"104": {
|
36 |
+
"content": "[UNK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
},
|
43 |
+
"30526": {
|
44 |
+
"content": "<mask>",
|
45 |
+
"lstrip": true,
|
46 |
+
"normalized": false,
|
47 |
+
"rstrip": false,
|
48 |
+
"single_word": false,
|
49 |
+
"special": true
|
50 |
+
}
|
51 |
+
},
|
52 |
+
"bos_token": "<s>",
|
53 |
+
"clean_up_tokenization_spaces": false,
|
54 |
+
"cls_token": "<s>",
|
55 |
+
"do_lower_case": true,
|
56 |
+
"eos_token": "</s>",
|
57 |
+
"extra_special_tokens": {},
|
58 |
+
"mask_token": "<mask>",
|
59 |
+
"model_max_length": 512,
|
60 |
+
"pad_token": "<pad>",
|
61 |
+
"sep_token": "</s>",
|
62 |
+
"strip_accents": null,
|
63 |
+
"tokenize_chinese_chars": true,
|
64 |
+
"tokenizer_class": "MPNetTokenizer",
|
65 |
+
"unk_token": "[UNK]"
|
66 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|